Offline reinforcement learning, a new method of reinforcement learning (RL) is about to make strides in AI technology. The method utilizes the previously collected data for training the algorithm and does not require interaction with the online environment. Richa Verma, a systems engineer at TCS research and PhD scholar at the RBCDSAI-IIT Madras, believes that offline RL has tremendous potential and can scale up systems in a huge way as these offline algorithms can be employed to automate decision-making in various domains like robotics, education and healthcare. Richa wishes to work on the current limitations of offline RL paving the path of its deployment in the real world.
Born and brought up in the nation’s capital - New Delhi, Richa was always good in academics and interested in science from early on. She loved Physics but refrained from taking up the subject for further studies as there were not many programmes in Physics that guaranteed employability. Richa, therefore, took up her second love, computer science, for her career and joined B.Tech in Computer Science programme in the New Delhi based Indraprastha University. During her graduation, she learned about artificial intelligence and got excited by its scope. To deepen her understanding of the subject, she decided to go for a master’s degree in Computer Science and gave the GATE exam for the same. She was among the top one percentile students that appeared for the GATE exam and got admitted to the master’s program in IIIT Delhi based on the GATE score. During her stay in IIIT Delhi, she was involved in various research projects. One of these projects entailed using geographical data to look for patterns of criminal activities in an area and the other involved building an animal detection module for a virtual assistant for visually impaired individuals in collaboration with IIT Delhi. She was also involved in developing an algorithm for a seismic sensor that detects the presence of elephants around it and can be used reduce elephant casualties on railway tracks. The research results of these projects were published in reputed conferences and journals.
Though Richa was always interested to go for a PhD, she decided to join the industry immediately after her master’s so as get an experience of real-world problems. Therefore, she joined TCS research as a research scientist in the year 2017 and has been involved in solving combinatorial optimization problems using reinforcement learning there since then. One of the projects she worked on involved developing an algorithm that makes loading plan on ports to involve minimizing the shuffling of containers on the ship. While working at the TCS research, she got more and more involved in theoretical aspects of reinforcement learning and attended several computer science conferences. She remembers meeting Prof. Balaraman Ravindran, professor at the IIT Madras, in these conferences who, along with Dr. Harshad Khadilkar (her mentor at TCS) motivated her to go for a PhD. Richa jumped at this opportunity and got enrolled for a PhD under his supervision. Fascinated by the scope of offline RL, she plans to work on the safety aspect of offline RL for her PhD work. She feels lucky to be part of RBCDSAI at IIT Madras and believes that such communities help in advancing science through brainstorming ideas and collaborations.
Inspired by the works of women scientists like Marie Curie, Dr. Fei Fei Li and Dr. Gagandeep Kang, Richa says that her mentors and peers at IIIT Delhi kickstarted her research journey. To motivate more girls towards science, she has been volunteering for women in data science events. When not working, she loves to read books and hike. Always eager to learn from her peers and mentors, Richa firmly believes in the quote “If you are the smartest person in the room, then you are in the wrong room.”